Posted
October 27, 2005

Self-Organizing Social Networks

The rise of social networks could have a profound impact on virtually every sector of society.

Not ten years ago, the accepted wisdom was that the Net was neither safe for significant transactions nor would people trust those whom they had not physically met. It was further assumed that the Net was the province of techies and that it would never become a consumer, global medium.

Today there are 500 million people on the Net, and as the Net becomes wireless and mobile, this figure will exceed several billion within the next decade. The United States Department of Defense is investing over $100 billion to develop and deploy a “global grid” to coordinate its worldwide military operations. Yahoo has over 130 million registered users, and eBay has over 120 million registered users and 30 million registered businesses.

One of the fastest-growing categories on the Net today is “social networks.” In just the last five years, social networking software – what some pundits herald as the first example of social technology software – has seen explosive growth. MySpace was nonexistent five years ago and now has over 31 million users. Yahoo has launched its own social networking platform, “360,” and Google has its version, Orkut.

Why the big fuss? Social networks are the beginnings of peoples around the word starting to grow social networks organizations and eventually institutions that mirror and express who they are and their interests. It is a new kind of online, digital zone where everyone can – to use Adam Smith’s words – “bicker, barter, and trade.” Social networks are less about place, or virtual place, than about trusted relationships and social context.

The term “social networks” is used very loosely and can be applied to a variety of phenomena – “social networking sites” such as Facebook, Friendster or MySpace, online gaming communities, e-commerce communities such as eBay and even open source communities. There is a close albeit vague correspondence between social networks and the notion of communities as used on the Net.

The advantage of treating online communities or groups as social networks is that it highlights the fact that different people are linked to one another. People can find and make friends, set up affinity groups, coordinate events and set up businesses. Through their interactions and behaviors, members of social networks are signaling their pleasures and displeasures. They are also leaving a digital record of how they treat one another. They are free to rate one another, and their likes and dislikes. They can even create their own information syndication services (RSS, for Really Simple Syndication), and share their playlists with other members of the social networking community.

In short, by making all sorts of micro-interactions visible as digital records, the Net may have given the social sciences a new kind of social lens, the equivalent of the van Leeuwenhoek microscope that it needs.

Online networks can be used to model social relationships. It becomes possible to measure the type and volume of relationships that people have by analyzing who they link to and who acts as nodes to many others. By charactering social interactions in this way, other insights become possible, such as the strength and frequency of social relationships, and the dependency of certain relationships upon one another. By analyzing the structure of the network, whether it is structured as a “hub and spoke” or a “random” network, it becomes possible to identity path dependencies, bottlenecks, dead-ends, outliers, connectors, and the opportunity to act as an intermediary among different social networks.

As useful as some as these insights are in treating social relationships as fixed networks, if one goes a step further and makes the network self-organizing, so that any node (i.e., any individual) can dynamically generate its own kind of links and assemble, recombine, and decouple sub-networks, then you have a very powerful means of understanding elemental aspects of human social interactions. You can begin to postulate the dynamics of basic social emotions such as trust, reciprocation, threats, punishment, reward, mutual support, innovation, and so forth. Self-organizing networks are a very flexible means of representation, and can also be used to map patterns of social signals that people use to communicate their intentions. These patterns, in turn, can be correlated to specific social emotions and neural activity.